History teaches us that when you're in the middle of sweeping technology-driven change, the struggle to simply keep pace can make it hard to maintain perspective. This is the situation we find ourselves in now, with the rapid emergence of artificial intelligence (AI). Suddenly there's great urgency to explore how businesses and society can realize the benefits of AI while simultaneously grappling with the wide range of issues it has raised--issues like the future of employment and how we can preserve individual privacy and protect public safety. To do all this in a smart and balanced way requires a long-term view of what the decisions we make now will mean for the future. This is something that is probably a little easier to do at a company like Schneider Electric.

State representation learning (SRL) in partially observable Markov decision processes has been studied to learn abstract features of data useful for robot control tasks. For SRL, acquiring domain-agnostic states is essential for achieving efficient imitation learning (IL). Without these states, IL is hampered by domain-dependent information useless for control. However, existing methods fail to remove such disturbances from the states when the data from experts and agents show large domain shifts. To overcome this issue, we propose a domain-adversarial and -conditional state space model (DAC-SSM) that enables control systems to obtain domain-agnostic and task- and dynamics-aware states. DAC-SSM jointly optimizes the state inference, observation reconstruction, forward dynamics, and reward models. To remove domain-dependent information from the states, the model is trained with domain discriminators in an adversarial manner, and the reconstruction is conditioned on domain labels. We experimentally evaluated the model predictive control performance via IL for continuous control of sparse reward tasks in simulators and compared it with the performance of the existing SRL method. The agents from DAC-SSM achieved performance comparable to experts and more than twice the baselines. We conclude domain-agnostic states are essential for IL that has large domain shifts and can be obtained using DAC-SSM.

In Machine Learning and Optimization community there are two main approaches for convex risk minimization problem. The first approach is Stochastic Averaging (SA) (online) and the second one is Stochastic Average Approximation (SAA) (Monte Carlo, Empirical Risk Minimization, offline) with proper regularization in non-strongly convex case. At the moment, it is known that both approaches are on average equivalent (up to a logarithmic factor) in terms of oracle complexity (required number of stochastic gradient evaluations). What is the situation with total complexity? The answer depends on specific problem. However, starting from work [Nemirovski et al. (2009)] it was generally accepted that SA is better than SAA. Nevertheless, in case of large-scale problems SA may ran out of memory problems since storing all data on one machine and organizing online access to it can be impossible without communications with other machines. SAA in contradistinction to SA allows parallel/distributed calculations. In this paper we show that SAA may outperform SA in the problem of calculating an estimation for population ({\mu}-entropy regularized) Wasserstein barycenter even for non-parallel (non-decenralized) set up.

American-style options are used not only by traditional asset managers but also by energy companies to hedge "optimised assets" by finding optimal decisions to optimise their P&L and find their value. A common modelling of a power plant unit P&L is done using swing options which are American options allowing to exercise at most l times the option with possibly a constraint on the delay between two exercise dates (see Carmona and Touzi (2008) or Warin (2012) for gas storage modelling). Formally, for T 0, we are given a stochastic processes ( X t) t 0 defined on a probability space (Ω, F, F ( F t) t 0, P) and one wants to find an increasing sequence of F stopping times τ ( τ 1,τ 2,...,τ l) that maximises the expectation of some objective function f E Pnull l null i 1f ( τ i,X τ i) 1 τ i Tnull . Numerical methods to solve the optimal stopping problem when l 1,f ( x,t) e rt g (x) and X is Markovian include: - Dynamic programming equation: the option price P 0 is computed using the following backward discrete scheme over a grid t 0 0 t 1 ... t N T: P t N g ( X T), P t i max( g ( X t i),e r (t i 1 t i) E P( P t i 1 F t i)), i 0,...,N 1 . One then needs to perform regression to compute the conditional expectations, see Longstaff and Schwartz (2001) or Bouchard and Warin (2012).

Joseph Bakarji 1, Daniel M. Tartakovsky 1 Department of Energy Resources Engineering, Stanford University, 367 Panama Mall, Stanford, 94305 CA, USAAbstract A general method for learning probability density function (PDF) equations based on Monte Carlo simulations of random fields is proposed. Sparse linear regression is used to discover the relevant terms of a partial differential equation of the distribution. The various properties of PDF equations, like smoothness and conservation, makes them very well adapted to equation learning methods. The results show a promising direction for data-driven discovery of coarse-grained equations in general. Introduction Probabilistic models have proven to be essential in various fields of science and technology for optimizing predictability under epistemic and model uncertainty.

Helsinki, Amsterdam, Copenhagen, Paris Region, Stavanger and Tallinn will challenge companies to develop energy and mobility solutions using artificial intelligence (AI) as well as 5G, Internet of Things (IoT) and other related technologies. The initiative is part of AI4Cities, a three-year EU-funded project bringing together European cities looking for AI solutions to reduce their greenhouse gas emissions and meet climate commitments. The cities and regions will go through a pre-commercial procurement (PCP) process, which allows them to steer the development of new solutions directly towards their needs. Once they have defined their requirements, the cities will challenge start-ups, SMEs and larger companies to design solutions applying the use of AI and other technologies. Total funding of €4.6 million will be divided among the selected suppliers throughout the whole PCP process.

Gaussian processes are ubiquitous in statistical analysis, machine learning and applied mathematics. They provide a flexible modelling framework for approximating functions, whilst simultaneously quantifying our uncertainty about this task in a computationally tractable manner. An important question is whether these approximations will be accurate, and if so how accurate, given our various modelling choices and the difficulty of the problem. This is of practical relevance, since the answer informs our choice of model and sampling distribution for a given application. Our paper provides novel approximation guarantees for Gaussian process models based on covariance functions with finite smoothness, such as the Mat\'ern and Wendland covariance functions. They are derived from a sampling inequality which facilitates a systematic approach to obtaining upper bounds on Sobolev norms in terms of properties of the design used to collect data. This approach allows us to refine some existing results which apply in the misspecified smoothness setting and which allow for adaptive selection of hyperparameters. However, the main novelty in this paper is that our results cover a wide range of observation models including interpolation, approximation with deterministic corruption and regression with Gaussian noise.

Despite the ubiquity of the Gaussian process regression model, few theoretical results are available that account for the fact that parameters of the covariance kernel typically need to be estimated from the dataset. This article provides one of the first theoretical analyses in the context of Gaussian process regression with a noiseless dataset. Specifically, we consider the scenario where the scale parameter of a Sobolev kernel (such as a Mat\'ern kernel) is estimated by maximum likelihood. We show that the maximum likelihood estimation of the scale parameter alone provides significant adaptation against misspecification of the Gaussian process model in the sense that the model can become "slowly" overconfident at worst, regardless of the difference between the smoothness of the data-generating function and that expected by the model. The analysis is based on a combination of techniques from nonparametric regression and scattered data interpolation. Empirical results are provided in support of the theoretical findings.

Recurrent neural networks (RNNs) are a powerful approach for time series prediction. However, their performance is strongly affected by their architecture and hyperparameter settings. The architecture optimization of RNNs is a time-consuming task, where the search space is typically a mixture of real, integer and categorical values. To allow for shrinking and expanding the size of the network, the representation of architectures often has a variable length. In this paper, we propose to tackle the architecture optimization problem with a variant of the Bayesian Optimization (BO) algorithm. To reduce the evaluation time of candidate architectures the Mean Absolute Error Random Sampling (MRS), a training-free method to estimate the network performance, is adopted as the objective function for BO. Also, we propose three fixed-length encoding schemes to cope with the variable-length architecture representation. The result is a new perspective on accurate and efficient design of RNNs, that we validate on three problems. Our findings show that 1) the BO algorithm can explore different network architectures using the proposed encoding schemes and successfully designs well-performing architectures, and 2) the optimization time is significantly reduced by using MRS, without compromising the performance as compared to the architectures obtained from the actual training procedure.

The raging Australian and Amazon wildfires have raised a burning question for all of us - why the very technology, that has been a major facilitator to human evolution and growth could not predict, manage or control its destruction? To those of us who are in the business of technology, it is time to ask a few tough questions in our boardroom meetings and take ownership of solving the problem. After all, what is growth worth if the planet itself is in peril? As someone who has witnessed the digital revolution unfold, I may not have a full-proof plan to address the climate emergency, in fact, we don't even have the visibility of all evolving technologies that may be required to solve the climate emergency. But, I am clear and convinced that we have to start now and start with the available technologies which in their own right are very powerful and transformational.